By Dr. Gary Deel
Faculty Director, School of Business, American Public University
This is the first of two articles on the use of Turnitin as a tool for teaching writing.
Many modern universities, including American Public University, use online learning management systems equipped with the Turnitin originality verification tool. Its purpose is to assess the originality of students’ work. But a big unanswered question is how this tool should be applied to assess student work.
Start a degree program at American Public University.
Students look to their institutions for guidance on what is acceptable for coursework. Due to a lack of discussion or consensus on this issue, classroom policies concerning what is acceptable have generally been left to the discretion of the instructors. When a university doesn’t provide a consistent framework for coursework, instructors must fill in the gaps with their own classroom rules and policies.
This can be confusing for students who may take five or more classes in a semester and dozens of classes over the course of their degree program. So the variability in originality standards among instructors opens the door to the notion that these prescriptions are purely arbitrary.
Higher Education Should Be Able to Reach an Accord on Assessing Originality in Student Work
One would think that universities — and perhaps higher education as a whole — should be able to reach an accord on some basic philosophies for assessing originality in student work.
First, it’s worth discussing how the Turnitin originality tool works to assess student submissions and determine their originality. When a student’s paper is submitted to Turnitin, the software compares the contents of the work against a variety of known pre-existing sources.
For example, Turnitin runs similarity audits against all work available on the web, including books, articles, websites, blogs and any other content that major search engines like Google can crawl. Turnitin also maintains a database of all student submissions that the program has ever assessed. This includes submissions by students in the same class or the same school, and submissions from students at different schools that also utilize Turnitin.
This results in a robust originality audit, but of course it isn’t foolproof. For example, what about works that are available only in print and not on the internet? And what about student submissions at universities that don’t use Turnitin?
Turnitin is used by over 15,000 institutions in 140 countries. But there are still some institutions that have not signed on to the service. Turnitin obviously cannot reach materials from those schools for comparison. Such are the inherent limitations of today’s originality verification efforts.
Turnitin Renders What It Calls an ‘Originality Verification (OV) Score’ from One to 100
Nonetheless, from its comparison to known sources Turnitin renders what it calls an “originality verification (OV) score” of one to 100. The score indicates the percentage of the paper that is a direct match to outside sources.
Turnitin obviously uses complex algorithms to parse which similarities are worthy of genuine scrutiny and which are not relevant. But even with Turnitin’s level of sophistication, instructors should not act on the OV score without further analysis. For this purpose, Turnitin includes settings that instructors can adjust for fine-tuning audits.
For example, there is a “bibliography exclusion” option that directs the software to ignore any content within a paper that follows after certain heading keywords, such as “References,” “Works Cited,” or “Bibliography.” Activating this setting usually makes sense when auditing papers, because references will be repeated across multiple works that cite the same sources.
These kinds of similarities, however, do not reflect the kind of unoriginal content that would be of concern in grading papers. So as an instructor, I always direct Turnitin to ignore the bibliographies.
Another tool allows instructors to direct Turnitin to ignore excerpts of content that flag a match to pre-existing sources, unless the match is a minimum number of words. The parameter can be set for any number of words from two up. This is helpful because Turnitin will occasionally flag matched content that is too short to reflect a genuine concern of borrowed or stolen content.
However, it’s worth noting that Turnitin estimates the chance of a 16-word excerpt being a coincidental match to a pre-existing source at less than one in a trillion. As such, I generally start my minimum-word-match tool at between eight and 10 words, and then adjust based on results from there.
One final tool allows instructors to tell Turnitin to ignore direct quotes; that is, any content found between quotation marks in the student’s submission. This particular tool has questionable utility, and requires an explanation of the difference between plagiarism and originality.
Failing to Cite an Original Work Constitutes Plagiarism
In higher education, we teach our students that whenever they use the work of others — whether it be their exact words or simply their thoughts and ideas — they must cite the work to give credit to the author. Failing to cite an original work constitutes plagiarism, and so we require our students to cite their sources.
However, citation aside, we also expect that the work our students produce will be predominantly original. That is, the vast majority of the work should be the student’s own thoughts and ideas; students should cite the work of other authors discussed only when necessary to lend credibility to the student’s work.
Even When a Paper Is Properly Cited, a Student’s Paper Can Still Be Unoriginal
Imagine an assignment for which the student quotes multiple outside authors and does so with impeccable citations, both within the text and at the end of the manuscript. But upon a Turnitin audit, we find that the OV score is 90 percent, meaning that quotes from other authors comprise 90 percent of the assignment, and just 10 percent is original content.
The student has delivered an assignment that is genuinely free of any plagiarism because he cited all sources where appropriate. However, the assignment is nonetheless patently unacceptable because the vast majority of it is unoriginal. The student essentially created a mosaic of quotes from other authors and presented almost nothing in terms of his own work.
This is where the Turnitin setting to ignore quotes becomes problematic. It may be helpful if an instructor is concerned only with plagiarism and not with originality. But if an originality assessment is what we’re after, then we would also want to consider quoted content. For this reason, I usually include such content in the audit.
What happens after we play with the Turnitin settings? In the second part, I’ll discuss some of the leg work that good instructors must do to ensure that the originality audit provides meaningful feedback.
About the Author
Dr. Gary Deel is a Faculty Director with the School of Business at American Public University. He holds a JD in Law and a Ph.D. in Hospitality/Business Management. He teaches human resources and employment law classes for American Public University, the University of Central Florida, Colorado State University and others.